Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Run end-to-end Playwright tests via the developer CLI.
.claude/skills/platformplatform-e2e/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 126% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 54% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 25% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -56% | 0% |
bashdotnet run --project developer-cli -- e2e [search-terms...] [--smoke] [--browser <name>] [--retries <n>] [--last-failed] [--only-changed] [--stop-on-first-failure] [--include-slow] [--self-contained-system <name>] [--no-wait-for-aspire] --quiet
Use developer-cli exactly as written - do not expand to an absolute worktree path.
search-terms - filter by test name, tag, or spec file--smoke - smoke suite only--browser <name> - chromium (default), firefox, webkit, safari, all--retries <n> - max retries for flaky tests--last-failed - re-run only the previous run's failures--only-changed - only spec files with uncommitted changes--stop-on-first-failure, -x - exit on first failure--include-slow - include @slow-tagged tests (excluded by default)--self-contained-system <name> - narrows to one SCS (e.g. main, account)--no-wait-for-aspire - skip the Aspire readiness check (use only when Aspire is already up)Aspire must be running. If not, start it via the aspire-restart skill first.
E2E runs are slow. Never blindly run the full suite to verify a small change.
--only-changed.--last-failed."user management"), a tag, or a spec file.--stop-on-first-failure (-x) when iterating so the run aborts on the first failure.--smoke, get those green, then run the rest.firefox / webkit after chromium passes - and skip chromium on those runs (--browser firefox).A flaky test run in isolation often passes - the bug only shows under parallel load. Reproduce with the full file or suite before declaring it fixed.
Without --quiet Playwright streams every step to the conversation. On success the CLI prints a single summary; on failure it prints the failed tests and where to find the report.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 2,862 | 4,389 | +53% | 1 | 1 | 0% | 394 | 889 | +126% | 0 | 0 | — |
case-02 | fail→pass | 3,547 | 4,557 | +28% | 1 | 1 | 0% | 537 | 828 | +54% | 0 | 0 | — |
case-03 | fail→fail | 7,024 | 6,517 | -7% | 1 | 1 | 0% | 1,097 | 933 | -15% | 0 | 0 | — |
case-13 | fail→pass | 4,146 | 1,741 | -58% | 1 | 1 | 0% | 652 | 818 | +25% | 0 | 0 | — |
case-04 | fail→pass | 7,480 | 2,207 | -70% | 1 | 1 | 0% | 1,231 | 914 | -26% | 0 | 0 | — |
case-05 | pass→pass | 6,734 | 1,930 | -71% | 1 | 1 | 0% | 1,018 | 861 | -15% | 0 | 0 | — |
case-06 | pass→pass | 5,074 | 2,386 | -53% | 1 | 1 | 0% | 817 | 820 | +0% | 0 | 0 | — |
case-07 | fail→fail | 6,100 | 5,891 | -3% | 1 | 1 | 0% | 1,052 | 1,093 | +4% | 0 | 0 | — |
case-08 | fail→pass | 12,754 | 2,121 | -83% | 1 | 1 | 0% | 1,999 | 881 | -56% | 0 | 0 | — |
case-09 | fail→pass | 8,130 | 2,798 | -66% | 1 | 1 | 0% | 1,321 | 774 | -41% | 0 | 0 | — |
case-10 | fail→pass | 13,141 | 4,051 | -69% | 1 | 1 | 0% | 2,054 | 1,233 | -40% | 0 | 0 | — |
case-11 | pass→pass | 14,645 | 2,177 | -85% | 1 | 1 | 0% | 2,172 | 821 | -62% | 0 | 0 | — |
case-12 | fail→pass | 8,229 | 4,437 | -46% | 1 | 1 | 0% | 1,246 | 814 | -35% | 0 | 0 | — |
case-14 | fail→fail | 7,792 | 4,938 | -37% | 1 | 1 | 0% | 1,094 | 806 | -26% | 0 | 0 | — |
case-15 | fail→fail | 5,697 | 4,948 | -13% | 1 | 1 | 0% | 821 | 806 | -2% | 0 | 0 | — |
case-16 | fail→pass | 1,949 | 1,569 | -19% | 1 | 1 | 0% | 245 | 787 | +221% | 0 | 0 | — |
case-17 | fail→pass | 7,723 | 1,941 | -75% | 1 | 1 | 0% | 1,135 | 866 | -24% | 0 | 0 | — |
case-22 | pass→pass | 5,561 | 4,100 | -26% | 1 | 1 | 0% | 938 | 1,264 | +35% | 0 | 0 | — |
case-18 | fail→pass | 7,976 | 1,883 | -76% | 1 | 1 | 0% | 1,414 | 830 | -41% | 0 | 0 | — |
case-19 | fail→fail | 6,435 | 4,941 | -23% | 1 | 1 | 0% | 1,068 | 886 | -17% | 0 | 0 | — |
case-20 | pass→pass | 4,714 | 1,342 | -72% | 1 | 1 | 0% | 709 | 750 | +6% | 0 | 0 | — |
case-21 | pass→pass | 5,329 | 2,521 | -53% | 1 | 1 | 0% | 909 | 950 | +5% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 17 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +50 percentage points is the difference between those two pass rates over the 17 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.